“The crux of the biscuit is the apostrophe.” –Frank Zappa
For those wanting all the details, take a look at my curriculum vitae(CV). To contact me decode this puzzle: ‘shawn <dot> garbett <at> vumc.org’ or ‘shawn <at> garbett.org’ for personal correspondence.
I did my undergraduate at Tulane University in New Orleans in electrical engineering and computer science. I worked for several years at TVA and applied combinatorial optimization to coal burning and linear programming to running the reservoir system. The program led to an ~2% increase in coal burn efficiency world wide for multiple furnace installations leading to massive savings–and lower power bills for all.
Worked as a consultant for many years, of which the capstone project was working for Air Force One through Walter Reed Hospital. It’s a great honor when one’s software is part of the Air Force One tooling–and the first closed loop medical device approved by the Food and Drug Administration.
I was hired hired full time in Vanderbilt’s Cancer Biology department to create math models of cancer progression and help run an automated microscope. Working with the microscope data, it led me to statistics as the best means to interpret the data for the lab. I discovered a new passion in statistics, and went back and got my masters at Penn State in 2014. This has led me to becoming faculty in Biostatistics at Vanderbilt University Medical Center. I work a fair bit with Health Policy on evaluating and optimizing medical treatment decision making. Our working group got a paper in the top 10 genomics papers of the year, three years in a row–an academic hat trick.
I am infinitely curious and highly resourceful. I love playing piano1 Preciso Me Encontrar performed by Buffy Rhea and myself., board games, modeling, reading sci fi, and programming. One theme through all these years is modeling–whether it be simulating nuclear magnetic resonance, playing an abstract simulation game or building a scale car and painting it.
Fluent in a wide variety of programming languages and technology stacks. However, what I really enjoy is formally typed languages like Haskell. It’s incredible that via the Curry-Howard correspondence each program is a witness to the proof of that program’s type. That type may be something utterly useless and broken, but one has never the less proved it.2 If we only knew what it was. Dependent type systems of proof via Homotopy Type Theory is a personal fascination.
If one needs predictive or explanatory statistical data science models3 Currently referred to as artificial intelligence in the press at present. of data; this I can deliver. Applied economic decision making with dimensions of cost and effectiveness is a specialty.
Send me your data!
These are a few of the packages I maintain.